If I understand the principle of neurotransmission correctly, then after binding to a receptor on the postsynaptic side, neurotransmitter molecules must be dissociated from the receptor and removed quickly from the synaptic cleft (see this book).
This article states that the off-rate $k_{\text{off}}$ for the 5HT3 receptor and some ligand is approximately $43 * 10^{-6} s^{-1}$. If we ignore the binding process for now and denote by $C$ the complex formed by 5HT3 and the ligand, we can describe the dissociation process by the following simple differential equation: $$ \frac{dC}{dt} = -k_{\text{off}} * C $$ (see Wikipedia). Using the $k_{\text{off}}$ from above, we get that the half time of this dissociation process is $\frac{\ln{2}}{k_{\text{off}}} \approx 16116s \approx 4,5h $. A similar $k_{\text{off}}$ rate for all other kinds of 5HT receptors binding serotonin can be derived from the dissociation constants found in the Ki Database.
But wouldn't that mean that, when serotonin is released from the presynaptic axon terminal and the 5HT3 receptor binds it, not even half of the neurotransmitter molecules are dissociated from the receptor after more than 4 hours? That seems to clearly contradict the statement from the beginning of this post - so is something wrong with my calculations or does it really take that long? Or have I understood something completely wrong about the principle of neurotransmission?